ISSN 2333-0694

JNAHThe Journal of North American Herpetology

Volume 2017(1): 28-33 29 March 2017 jnah.cnah.org

HABITAT UTILIZATION BY THE TEXAS HORNED (PHRYNOSOMA CORNUTUM) FROM TWO SITES IN CENTRAL TEXAS WESLEY M. ANDERSON1,2,6, DAVID B. WESTER1,3, CHRISTOPHER J. SALICE4,5, AND GAD PERRY1

1Department of Natural Resources Management, Texas Tech University, Lubbock, TX 79409, USA 2Present address: Department of Wildlife Ecology and Conservation, University of Florida, Gainesville, FL 32611, USA 3Present address: Caesar Kleberg Wildlife Research Institute and Department of , Rangeland, and Wildlife Sciences, Texas A&M University-Kingsville, Kingsville, TX 78363, USA 4The Institute of Environmental and Human Health, Texas Tech University, Lubbock, TX 79409, USA 5Present address: Department of Biological Sciences, Towson University, Towson, MD 21252, USA 6Corresponding author: [email protected]

ABSTRACT — The Texas (Phrynosoma cornutum) is found in a variety of habitats. Although several studies have been conducted on habitat use by this , none have been performed in central Texas, a more mesic habitat than most of those previously studied. This area is of special interest because horned lizard populations have been experiencing sharp declines in central Texas over the last approximately 50 years. We collected habitat data at two sites in central Texas, Camp Bowie and Blue Mountain Peak Ranch. Microhabitat data included canopy cover and ground cover from digitized photographs of Daubenmire quadrats; macrohabitat variables included vegetation height and length, cactus height, soil penetrability, woody plant species richness, tree density, tree diameter at breast height (DBH), and density of mounds collected along 100-m by 2-m transects. Similar patterns of habitat use were observed between the two sites. At Blue Mountain Peak Ranch, appeared to be located in areas with a diversity of ground cover types, as observed in previous studies. At Camp Bowie, vegetation encroachment limited lizards in some areas to the use of roads and road margins. Implementation of prescribed burns or other vegetation management could create the preferred ground cover mosaic at such sites.

INTRODUCTION in coastal North Carolina, South Carolina, and Florida The genus Phrynosoma, collectively known as horned (Palmer and Braswell 1995). lizards, contains some of the most distinct examples of Unfortunately, P. cornutum has declined over much North American saurian fauna – both in terms of general of its geographic range. Over the last several decades appearance and aspects of their biology. Phrynosoma populations have declined precipitously in central and species are characterized by a dorsoventrally flattened east Texas and parts of Oklahoma (Carpenter et al. 1993, body, occipital horns, and a primarily myrmecophagous Donaldson et al. 1994; Henke 2003). In Texas the species diet. Other notable characteristics include late maturation has a state designation of threatened (Donaldson et al. – many species do not breed until their second year – 1994). It is also listed as a species of special concern and unusually large relative clutch masses (Pianka and in Colorado and Oklahoma. Hypotheses for the causal Parker 1975). factors for these declines include effects of pesticides, The (Phrynosoma cornutum) impact of the red imported fire ant Solenopsis( invicta), is found throughout much of the south-central and habitat loss, and the legacy of collection for the pet or southwestern United States, as well as the northern curio trade (Donaldson et al. 1994). states of Mexico. Throughout this range it occupies a Understanding an organism’s habitat requirements is variety of habitats including shortgrass prairie, mesquite- essential in making informed management decisions. grassland, shrubland, oak savanna, and desert (Price Studies of habitat preferences for P. cornutum have 1990). It has also been transported throughout the been conducted in several localities throughout its range country, largely through intentional collecting by would- including south Texas (Fair 1995; Burrow et al. 2001), be pet keepers, and breeding populations have become northwest Texas (Henry 2009), and Colorado (Montgomery established outside of its native range including areas and Mackessy 2003). Studies focused primarily on

© The Center for North American Herpetology 28 habitat use in relation to habitat management, such as to manage plant communities and preserve biodiversity, prescribed burns or grazing have also been conducted, but these activities were mostly suspended during our mostly in Texas (Fair and Henke 1997; Hellgren et al. data collection due to training activities. Blue Mountain 2010; Inslee 2010) but also in Kansas (Wilgers and Peak Ranch is managed primarily for habitat restoration Horne 2006). However, no research on habitat use for in an effort to increase water to the aquifer and increase this species has been conducted in central Texas, where it species diversity on the property. An aggressive has experienced the most dramatic population declines. management plan has been implemented – including Our objective was to provide managers and restoration regular prescribed fires and brush removal – primarily to ecologists with quantitative habitat use data based on reduce Ashe juniper (Juniperus ashei) density. current lizard encounters. We documented habitat use at Methods — We located individual P. cornutum either two spatial scales: the microhabitat (immediately near by conducting surveys via road cruising at 20 km/h each animal) level and the macrohabitat (patch) level. or less or through fortuitous encounters; lizards were captured by hand. Upon capture we recorded date, time MATERIALS AND METHODS of day, GPS coordinates (UTM, Garmin GPSmap 76CS), Study Area — This study was conducted at two sites and sex. We determined the sex by examining the located 128 km apart in central Texas (Fig. 1). From cloacal region to check for the presence of a hemipenal April 2007 to October 2010 we collected data at Camp bulge and/or femoral pores. Individuals with either of Bowie, a 3,542 ha Texas National Army Guard (TXANG) these characteristics present were recorded as males. training facility located 7 km southeast of Brownwood, in Hatchlings or juveniles that were too small to exhibit these Brown County, Texas, and from June 2008 to November characteristics or those individuals whose sex we could 2010 at Blue Mountain Peak Ranch, a 335 ha private not confidently discern were assigned as “undetermined.” ranch located 29 km southwest of Mason, in Mason Snout-vent length (SVL) and tail length were measured County, Texas. Camp Bowie is managed primarily for with a ruler to the nearest mm and mass was taken training military personnel, although limited hunting using an appropriately-sized spring scale (micro-line and fishing have at times been allowed by permit. Camp spring, PESOLA AG, Baar, Switzerland) to the nearest Bowie has a natural resources management plan, which 0.1 g. We permanently marked individuals larger than includes the use of prescribed fire and brush removal 20 g and 65 mm SVL with a PIT tag (passive integrated

Camp Bowie

Blue Mountain Peak Ranch

Figure 1. Aerial photographs of study sites and associated site boundaries in yellow and red for Blue Mountain Peak Ranch and Camp Bowie, respectively.

Journal of North American Herpetology 2017(1) 29 Table 1. Microhabitat ground cover by treatment and encounter type at Blue Mountain Peak Ranch. Each cover type was analyzed with a test of equality of binomial proportions with α = 0.05. Location: site where a lizard was initially seen. Because many lizards were first seen on a road, data for sightings that could be more representative are provided separately. Initial Encounter for All Lizards Initial Off-Road Encounter Ground Location Random Location Random Cover Type (% ± SD) (% ± SD) p-value (% ± SD) (% ± SD) p-value

Bare Ground 43.06 ± 23.28 25.44 ± 17.05 < 0.0001 26.89 ± 16.29 30.63 ± 17.94 0.0109 Litter 4.94 ± 8.87 13.50 ± 19.20 < 0.0001 9.79 ± 10.93 11.84 ± 18.26 0.0415 Grass 17.76 ± 19.95 35.00 ± 21.57 < 0.0001 31.63 ± 25.15 24.89 ± 20.37 < 0.0001 Forb 5.00 ± 15.55 5.59 ± 6.29 0.2786 8.37 ± 20.40 5.32 ± 5.41 0.0002 Cactus 0.00 ± 0.00 0.76 ± 4.12 < 0.0001 1.63 ± 7.11 0.26 ± 1.15 < 0.0001 Wood 1.62 ± 5.57 1.97 ± 5.14 0.2726 6.26 ± 12.43 2.79 ± 5.26 < 0.0001 Rock 4.24 ± 7.23 1.97 ± 4.22 < 0.0001 1.05 ± 1.65 0.84 ± 1.57 0.5025 Cobble 1.32 ± 2.28 4.29 ± 7.26 < 0.0001 2.68 ± 4.01 5.58 ± 6.30 < 0.0001 Pebble 22.03 ± 19.68 11.47 ± 10.68 < 0.0001 11.68 ± 11.87 17.84 ± 14.77 < 0.0001 Manmade 0.03 ± 0.17 0.00 ± 0.00 0.2390 0.00 ± 0.00 0.00 ± 0.00 1.0000 transponder tag, ID 100A, Trovan, UK) by implanting it R-1000, Communications Specialists, Inc.) in conjunction subcutaneously on the dorsum slightly anterior to the with a 3-element folded Yagi antenna. right hip with a syringe and 12-gauge stainless steel During the 2009 and 2010 field seasons, we collected needle (Camper and Dixon 1988). Incisions were closed microhabitat data. Upon initial sighting for each lizard (n with superglue. Individuals too small to be PIT tagged = 34), a 0.5 m x 0.5 m (0.25 m2) quadrat constructed of were temporarily marked with a unique code using a rebar was centered at the point of observation with one permanent marker on the ventral side. side oriented north. A photograph of the quadrat was Between 2007 and 2010 lizards from both sites were taken at a constant height of 1 m with a 9.0 megapixel fitted with radio transmitters (Wildlife Materials SOPB- digital camera (PowerShot SX110 IS, Canon). Using 2070: 2007-2009, Holohil BD-2: 2008-2009, and ATS image analysis software (SamplePoint v. 1.51, USDA R1645/R1655: 2010; Fig. 2). Seventeen lizards (seven ARS), we calculated crown cover percentages of bare males and ten females) and nineteen lizards (nine males ground, litter (dead organic matter), grass, forb, cactus, and ten females) were telemetered from Camp Bowie embedded rock (> 4 cm at widest point), cobble (> 4 and Blue Mountain Peak Ranch, respectively. Transmitters cm at widest point), pebble (< 4 cm at widest point), were attached between the shoulder blades using clear and manmade objects. To determine relative availability silicone. To avoid loss of the transmitter when the lizard of these habitat characteristics, each lizard relocation molts, we secured a zip tie collar around the lizard’s neck point was paired with a point 10 m away in a random and attached the transmitter to this collar with fishing direction selected from a random number table. Data for line. In order to better camouflage the telemetered this random location were collected in the same manner lizard, additional silicone was spread around the as described above. remaining exposed area of the transmitter and covered All telemetered lizards that were initially captured on with native soil. The average weight for the transmitter roads were relocated approximately one week after assembly from 2009 and 2010 was 6.6 ± 2.4% (range - release (n = 19), at which point the same habitat 3.2-13.1%) of the lizard’s body weight at time of capture measurements were taken. This was in an effort to (Knapp and Abarca 2009; Mougey 2009). We relocated see what impact collecting data on roads might have lizards using a hand-held receiver (Telemetry Receiver on microhabitat variables and allow us to draw better conclusions by understanding any potential biases. Because only three lizards were located at Camp Bowie in 2009 and 2010, and two of those were only found while surveying on a road, microhabitat use for Camp Bowie was not quantified. A pocket penetrometer (pocket penetrometer E-280, Geotest Instrument Corp.) was used to measure soil penetrability (kg/cm2) during both years. Measurements were taken in the northeast, southeast, southwest, and northwest corners of the plot. Penetrability measurements for each corner were averaged and a penetrability measurement for the quadrat was calculated. As the penetrometer could not measure any surface with a penetrability > 4.50 kg/cm2, these measurements were recorded as “> 4.50” and excluded from the quadrat average. For each quadrat we then tabulated how many of the four penetrability measurements were greater than 4.50 kg/cm2. In order to examine habitat use at a larger (patch) spatial scale, we established 100 m x 2 m belt transects. At both Camp Bowie and Blue Mountain Peak Ranch, Figure 2. Two Texas Horned Lizards (Phrynosoma cornutum) we measured habitat variables along 30 transects encountered during this study at Blue Mountain Peak Ranch. The originating at locations where an individual lizard was individual on the right has been equipped with a radio transmitter. first observed and 30 transects originating at random

© The Center for North American Herpetology 30 Table 2. Transect habitat characteristics (mean ± 1 standard deviation) by treatment and site. Each habitat characteristic was analyzed with a Wilcoxon Signed Rank Test with α = 0.05. Camp Bowie Blue Mountain Peak Ranch Ground Cover Type Location Random p-value Location Random p-value

Herbaceous vegetation length (mm) 622.84 ± 201.55 567.20 ± 155.36 0.0412 576.57 ± 122.91 526.20 ± 141.06 0.0992 Herbaceous vegetation height (mm) 526.19 ± 180.17 455.90 ± 134.89 0.0173 477.70 ± 115.50 416.87 ± 137.77 0.0565 Cactus height (mm) 4.612 ± 18.05 5.10 ± 14.44 0.4532 5.03 ± 15.77 0.00 ± 0.00 0.0815 Harvester ant nest density (nests/transect) 0.23 ± 0.62 0.03 ± 0.18 0.0964 0.67 ± 0.25 0.67 ± 0.25 1.0000 Red imported fire ant nest density (nests/transect) 0.81 ± 1.38 1.43 ± 1.70 0.0769 0.73 ± 1.01 1.20 ± 1.35 0.1292 Soil penetrability (kg/cm2) 2.55 ± 0.57 2.57 ± 0.56 0.7345 1.43 ± 0.82 1.65 ± 0.44 0.0122 Woody plant richness (species/transect) 5.06 ± 3.88 9.33 ± 4.28 0.0003 8.60 ± 3.11 8.30 ± 2.58 0.6074 Tree density (trees/transect) 1.87 ± 3.32 4.70 ± 4.06 0.0004 2.03 ± 2.61 2.83 ± 4.88 0.6787 Tree diameter at breast height (DBH; cm) 7.71 ± 11.71 11.57 ± 7.53 0.0177 9.39 ± 9.22 11.66 ± 11.94 0.4403

points. Location transects were established based on the was different. Whereas initial lizard encounters were in 30 most recently located individual lizards at a site in an areas of significantly higher bare ground and pebble cover effort to avoid changes in habitat characteristics as much and significantly lower grass cover, initial off-road lizard as possible. All transects were oriented in a random encounters were at locations significantly higher in grass direction from the initial point. If a transect intersected a cover and lower in bare ground and pebble cover. Although property line before it reached 100 m, the property line bare ground and herbaceous vegetation accounted became 0 m and the transect proceeded 180º from the for over 60% of average ground cover for individuals original direction. at Blue Mountain Peak Ranch, woody vegetation was Habitat variables were recorded within a 0.5 m x 0.5 poorly represented. Although telemetered lizards were m frame as described above, but five photographs sometimes observed thermoregulating within the shade were taken (at 20 m intervals starting at 20 m) along of woody vegetation during midday, they were much the centerline of each transect. Within each quadrat we more frequently observed under the fringe of a grass measured greatest herbaceous vegetation height and clump or under cobble during this time. Lizards were length (distance from the base of the stem to tip) to the observed partially concealed in this manner during 173 nearest 1.0 mm, greatest cactus height to the nearest of 409 (42.3%) observations at Blue Mountain Peak 1.0 mm, and soil penetrability. Additional measurements Ranch. During the heat of the day, telemetered lizards within each belt transect included number of woody plant frequently would be located underneath the same rock species, number and diameter at breast height (DBH) of they were last observed underneath (pers. obs). Lizards all trees along transect (species of woody vegetation > were observed fully concealed or buried during 92 of 409 3 m in height and typically possessing a single trunk), (22.5%) observations. number of red imported fire ant mounds (at least 51% of At Camp Bowie, habitat along transects where lizards mound within one meter of centerline of transect), and had been observed was characterized by greater number of (Pogonomyrmex barbatus) herbaceous vegetation length, greater herbaceous mounds (at least 51% of mound within one meter of vegetation height, presence of fewer woody plant centerline of transect). species and fewer trees compared to transects at random All SamplePoint data for both microhabitat and transect locations (Table 2). Soil at Blue Mountain Peak Ranch quadrats were analyzed using a test of equality of was more penetrable at locations utilized by lizards than

binomial proportions. This test considers the individual at random locations (Z59 = 2.506, p = 0.012) and was point in each quadrat as the sampling unit and is the only habitat characteristic that differed between preferable to a chi-square test where many of the cover location and random transects. The mean number of S. type categories would have been combined to meet invicta nests did not significantly differ between lizard assumptions inherent to that test. All tests of equality and random sites at either Camp Bowie or Blue Mountain of binomial proportions were performed in Program R v. Peak Ranch (Table 2). 2.10.1. All other data were analyzed in JMP v. 8.0. Log- Grass, bare ground, litter, and pebble were the most transformations were ineffective at stabilizing residual common patch-level ground cover categories at both variance or reducing skew; therefore, all analyses were Camp Bowie and Blue Mountain Peak Ranch (Table 3). All conducted using a Wilcoxon Signed Rank Test. All tests ground cover categories at Camp Bowie differed between were considered significant at theα = 0.05 level. lizard locations and random locations. Results at Blue Mountain Peak Ranch were similar. Canopy cover in areas RESULTS utilized by lizards was significantly lower than on random Microhabitat ground cover data for lizards at Blue transects at both Camp Bowie (p < 0.0001) and Blue Mountain Peak Ranch differed between initial encounter Mountain Peak Ranch (p = 0.0038; Table 3). locations and data collected off-road (Table 1). For initial lizard location and off-road location cover data, the DISCUSSION three most common cover categories were the same: Our findings match those of studies in northwest Texas bare ground, grass, and pebble. Habitat use versus (Henry 2009) and south Texas (Burrow et al. 2001), availability between the two encounter types; however, which indicate that lizards select areas containing

Journal of North American Herpetology 2017(1) 31 Table 3. Transect ground cover type and canopy cover by treatment and site. Each cover type was analyzed with a test of equality of binomial proportions with α = 0.05. Camp Bowie Blue Mountain Peak Ranch Location Random Location Random (% ± SD) (% ± SD) p-value (% ± SD) (% ± SD) p-value

Ground Cover Type Bare Ground 21.57 ± 13.63 17.57 ± 9.70 < 0.0001 16.16 ± 8.83 17.7 ± 5.85 0.0004 Litter 10.12 ± 8.69 14.63 ± 14.28 < 0.0001 11.91 ± 7.97 10.86 ± 6.34 0.0043 Grass 57.15 ± 19.94 54.06 ± 16.52 < 0.0001 51.18 ± 12.26 47.41 ± 13.14 < 0.0001 Forb 2.43 ± 2.39 4.00 ± 5.31 < 0.0001 2.05 ± 1.84 2.33 ± 1.83 0.1057 Cactus 0.07 ± 0.36 0.26 ± 0.82 < 0.0001 0.33 ± 1.47 0.00 ± 0.00 < 0.0001 Wood 3.06 ± 4.95 5.27 ± 5.59 < 0.0001 4.88 ± 6.21 6.53 ± 7.77 < 0.0001 Rock 0.22 ± 0.48 0.47 ± 1.17 0.0001 1.81 ± 2.13 3.01 ± 3.82 < 0.0001 Cobble 0.20 ± 0.55 1.25 ± 3.89 < 0.0001 4.14 ± 4.01 4.71 ± 4.89 0.0158 Pebble 5.17 ± 7.77 2.49 ± 3.68 < 0.0001 7.53 ± 6.07 7.45 ± 5.69 0.8093 Manmade 0.00 ± 0.00 0.00 ± 0.00 1.0000 0.01 ± 0.04 0.00 ± 0.00 0.2390

Canopy Cover Canopy 9.27 ± 13.00 20.38 ± 15.29 <0.0001 9.19 ± 9.59 10.17 ± 11.21 0.0038 a mosaic of cover types – particularly bare ground, counted through transect measurements, it may be that herbaceous vegetation, and woody vegetation. Given there was not adequate ant mound density to test for the geology of central Texas, particularly the Edwards an effect. Large S. invicta colonies may contain 220,000 Plateau, the numerous cobbles and boulders at our sites workers (Tschinkel 1988), thus even with a small may serve a thermoregulatory role usually provided increase in mound density, worker density could increase by woody vegetation elsewhere within P. cornutum’s substantially. The relationship between lizard and fire ant range because a lizard could bask on top of, hide in the density requires additional study. A more effective way shadow of, or hide completely underneath a suitably- to examine this relationship may be to quantify foraging sized rock. Contrary to our findings, Henry (2009) worker ant density through the use of bait traps. stated that, outside of areas around human habitation, Whereas most of Blue Mountain Peak Ranch offered lizards were very rarely seen using any sort of rock for suitable habitat, relatively little of Camp Bowie did. The thermoregulation or refugia. assertion that Blue Mountain Peak Ranch contained larger At Camp Bowie, herbaceous vegetation tended to amounts of suitable habitat is supported by the apparent be taller and tree and shrub densities lower in areas higher density of lizards (at least in 2009 and 2010) where lizards had been observed as opposed to random compared to Camp Bowie (Anderson 2012). The mosaic locations. This suggests lizards may be avoiding areas of bare ground and grass reported by Burrow et al. of higher woody vegetation density. Indeed, lizards (2001) was not present at Camp Bowie as it was at Blue were found in areas that, on average, had half as much Mountain Peak Ranch. Although percent ground cover canopy cover compared to cover at random locations composition along transects in known lizard locations (9.3% vs 20.4%). Interestingly, canopy cover in areas was similar between the two sites, the distribution was inhabited by lizards at Camp Bowie and at Blue Mountain very different. Whereas there existed a ground cover Peak Ranch were very similar at about 9%, cover which mosaic at Blue Mountain Peak Ranch, at Camp Bowie may be ideal for P. cornutum in this part of its range. most of the bare soil was confined to the roads and the At Blue Mountain Peak Ranch, the only habitat feature grass grew densely next to the roads. that differed between areas used by lizards and random Encounter observations and relocation of telemetered locations was soil penetrability. Soils in areas where individuals at Camp Bowie suggest the lizards were lizards had been encountered tended to be significantly largely confined to the roads with only marginal use of more penetrable than average. Fair (1995) similarly the adjacent fields. Newbold (2005) found that invasive found that in south Texas lizards prefer sandy loam soils grass significantly lowered the mobility of Desert that are easily penetrable for activities such as burrowing Horned Lizards (P. platyrhinos). Because of their wide, for thermoregulation, crypsis, and nesting. At Blue dorsoventrally flattened body, it is likely that horned Mountain Peak Ranch the predominant soil type was the lizards in general have a difficult time moving through Tarrant-Rock outcrop complex (National Cooperative Soil any species of dense grass. This may be the reason P. Survey 2013A), which is characterized by 10% surface cornutum at Camp Bowie, where habitat management cover of cobbles, stones, or boulders and very cobbly was less intensive and vegetation was denser, appeared clay to 33 cm below ground surface (bgs). Whereas the confined to roads. Road utilization by lizards puts them two predominant soil types at Camp Bowie (Douldle-Real into direct danger from human-induced mortality since association and Real association) tended to have fewer roads at Camp Bowie are frequently traversed by military rocks at the surface than the soils at Blue Mountain Peak vehicles. Similarly, the homogeneity of the road surface Ranch (National Cooperative Soil Survey 2013B), our likely reduces their ability to effectively employ crypsis studies indicate that these cobbly, silty clay, and very – this species’ primary defense – to avoid detection by gravelly loams tended to be on average more compacted. predators. Habitat for P. cornutum at Camp Bowie could Invasion by S. invicta has been indicated as a potential be improved through the use of prescribed burning, cause of horned lizard declines (Donaldson et al. 1994). mechanical or chemical woody vegetation control, We were unable to show a difference in ant mound carefully-managed grazing, or a combination of these densities between lizard sites and random locations. strategies. However, because of the relatively low numbers of nests Overall, the outlook for P. cornutum throughout Texas

© The Center for North American Herpetology 32 and elsewhere remains uncertain. Although research Hellgren, E.C., A.L. Burrow, R.T. Kazmaier, and D.C. from across its range continues to increase our knowledge Ruthven III. 2010. The effects of winter burning and of aspects of its biology and responses to habitat grazing on resources and survival of Texas horned management, important research questions remain lizards in a thornscrub ecosystem. The Journal of unexamined. Researchers should focus their efforts on Wildlife Management 74(2):300-309. testing the hypotheses related to this species’ decline. Henke, S.E. 2003. Baseline Survey of Texas Horned Data concerning egg and juvenile survival rates and Lizards, Phrynosoma cornutum, in Texas. The cause-specific mortality as well as the effects of fire Southwestern Naturalist 48(2):278-282. on this species would be particularly valuable (Anderson Henry, E. 2009. Ecology, home range, and habitat 2012). Until these data and a better understanding of selection of the Texas horned lizard (Phrynosoma factors behind its decline are available, scientists and cornutum) in the southern high plains of Texas. natural resource managers should continue managing Unpublished Master’s thesis, Texas Tech University, habitat to meet this species’ habitat preferences. Lubbock, TX. Acknowledgements — We thank R. Gundy, J. Forbus, Inslee, A.A. 2010. Herpetofaunal response to prescribed and J. Brewer for assistance in the field. We also thank R. burning on Matagorda Island, Texas: with emphasis Taylor for access to Blue Mountain Peak Ranch and for his on Texas horned lizards. Unpublished Master’s thesis, hospitality. Staff Sgt. N. Crownover and other personnel Fort Hays State University, Hays, KS. at Camp Bowie helped coordinate access to the training Knapp, C.R. and J.G. Abarca. 2009. Effects of radio base. This study was funded by a grant from the Texas transmitter burdening on locomotor ability and survival Army National Guard. Work was conducted under permits of iguana hatchlings. Herpetologica 65(4):363-372. from Texas Parks and Wildlife Department (0390-030) Montgomery, C.E. and S.P. MacKessy. 2003. Natural and the Texas Tech University IACUC (07038-07). This history of the Texas horned lizard, Phrynosoma is manuscript T-9-1247 of the College of Agricultural cornutum (), in southeastern Sciences and Natural Resources, Texas Tech University. Colorado. Southwestern Naturalist 48(1):111-118. Mougey, K. 2009. Radio transmitter mass: Impacts LITERATURE CITED on home range, daily displacement, and endurance Anderson, W.M. 2012. Home range, habitat use, and in Texas horned lizards and bearded dragons. conservation of the Texas horned lizard in central Unpublished Master’s thesis. Texas Tech University, Texas. Unpublished Master’s thesis, Texas Tech Lubbock, Texas. University, Lubbock, TX. National Cooperative Soil Survey. 2013A. Custom Soil Burrow, A.L., R.T. Kazmaier, E.C. Hellgren, and D.C. Resource Report for Mason County, Texas. Accessed Ruthven, III. 2001. Microhabitat selection by Texas on August 12, 2013: http://soils.usda.gov. horned lizards in southern Texas. The Journal of National Cooperative Soil Survey. 2013B. Custom Soil Wildlife Management 65(4):645-652. Resource Report for Brown and Mills Counties, Texas. Camper, J.D. and J.R. Dixon. 1988. Evaluation of a Accessed on August 12, 2013: http://soils.usda.gov. microchip marking system for amphibians and Newbold, T.A.S. 2005. Desert horned lizard (Phrynosoma . Texas Parks and Wildlife Report. 7100-159:1- platyrhinos) locomotor performance: the influence of 22. cheatgrass (Bromus tectorum). The Southwestern Carpenter, C.C., R. St. Clair, and P. Gier. 1993. Naturalist 50(1):17-23. Determination of the distribution and abundance of Palmer, W.M. and A.L. Braswell. 1995. Reptiles of North the Texas horned lizard (Phrynosoma cornutum) in Carolina. The University of North Carolina Press: Oklahoma. Final report. Federal Aid Project E-18, Chapel Hill. Oklahoma Department of Wildlife Conservation, Pianka, E.R. and W.S. Parker. 1975. Ecology of horned Oklahoma City, Oklahoma, U.S.A. lizards: a review with special reference to Phrynosoma Donaldson, W., A.H. Price, and J. Morse. 1994. The platyrhinos. Copeia 1975:141-62. current status and future prospects of the Texas Price, A.H. 1990. Phrynosoma cornutum. Catalogue of horned lizard (Phrynosoma cornutum) in Texas. Texas American Amphibians and Reptiles. 469: 1-7. Journal of Science 46:97-113. Tschinkel, W.R. 1988. Colony growth and the ontogeny Fair, W.S. 1995. Habitat requirements and capture of worker polymorphism in the fire ant, Solenopsis techniques of Texas horned lizards in south Texas. invicta. Behavioral Ecology and Sociobiology 22:103- Unpublished Master’s thesis. Texas A&M University – 115. Kingsville. Kingville, TX. Wilgers, D.J. and E.A. Horne. 2006. Effects of different Fair, W.S. and S.E. Henke. 1997. Effects of habitat burn regimes on tallgrass prairie herpetofaunal manipulations on Texas horned lizards and their prey. species diversity and community composition in the Journal of Wildlife Management 61:135-138. Flint Hills, Kansas. Journal of Herpetology 40:73-84.

Journal of North American Herpetology 2017(1) 33